Bi-amping and Audiolense/REQ

[Olli]

I am currently using Audiolense as a REQ software for integrating subs with mains.
I am using Roon‘s convolution engine for applying the generated filters.

I have booksheld 3 way speakers TAD CE-1 with bi-wiring terminals.

Question:

Can I use 2 different amps (eg a Nagra Classic Amp for mid and high range and a Class D amp (Hypex based or similar) for the lows) with the bi-wiring terminals and a MC DAC copying the XO parameters of the passive XO of the bookshelf speakers into Audiolense and thus creating a virtual 3 way system in Audiolense? I do not want to disconnect the internal passive XO of the TADs for now...

 

[solderdude]

I would just bi-amp them. Thus not do a double crossover.

The whole point of active crossover is to get rid of the passive parts.

 

[Olli]

Agree, in principle. However, I am aiming at something different here: I want to time allign all drivers in my system. the EQ software Audiolense does exactly that.

So if I biamp the TADs I can have 2 channels for the subs below 70 Hz or so, and direct 4 more channels to the Bi amping terminals of the TADs.
There's one 18cm cone Woofer and one coaxial Midrange/tweeter unit; so I could time align these three drivers with Audiolense when bi amping.

The XO frequewncies are 250 Hz and 2 Khz; I would correspondingly apply digital XOs in Audiolense at <70 Hz or so for the Subs, >70Hz<250 Hz for the TAD woofer and <250 Hz for the coaxial Midrange/tweeter unit.

My main question is if using 2 different amps for bi-amping the TADs, e.g.

1) a March Audio 502 for the woofers:

Power Output
2 Ohms - 450 W rms
4 Ohms - 500 W rms
8 Ohms - 350 W rms
Power Output Bridged
8 Ohms - 1.2 kW rms
Current Output - 27 A

THD + N - 0.0018% (Pout/2)
Output Noise - 40uV
Signal To Noise Ratio - 124 dB
Output Impedance - 2.6 mOhms
Frequency Response - 10Hz to 50kHz (+0/-3dB)
Voltage Gain 26dB

2) and this one (Nagra Classic Amp) for the coaxial Midrange/tweeter unit

Power 100 W RMS per channel into 8 Ω Sensitivity 1 V or 2 V RMS Bandwidth 10 Hz to 80 KHz, +0/-3 dB Crosstalk >70 dB Signal-to-noise ratio Typically 110 dB ASA “A” weighted THD+N < 0.05% Input impedance >100 KΩ Automatic start For input level >10 mV Protection, Overheating > +60° C (140° F) deactivates the amplifier DC speaker protection above ± 2.5 VDC

could do any harm to the speakers or could cause any other problem.

Thanks!

 

[March Audio]

Agree, in principle. However, I am aiming at something different here: I want to tune allign all drivers in my system. the EQ software Audiolense does exactly that.

So if I biamp the TADs I can have 2 channels for the subs below 70 Hz or so, and direct 4 more channels to the Bi amping terminals of the TADs.
There's one 18cm cone Woofer and one coaxial Midrange/tweeter unit; so I could time align these three drivers with Audiolense when bi amping.

The XO frequewncies are 250 Hz and 2 Khz; I would correspondingly apply digital XOs in Audiolense at <70 Hz or so for the Subs, >70Hz<250 Hz for the TAD woofer and <250 Hz for the coaxial Midrange/tweeter unit.

My main question is if using 2 different amps for bi-amping the TADs, e.g.

1) a March Audio 502 for the woofers:

Power Output
2 Ohms - 450 W rms
4 Ohms - 500 W rms
8 Ohms - 350 W rms
Power Output Bridged
8 Ohms - 1.2 kW rms
Current Output - 27 A

THD + N - 0.0018% (Pout/2)
Output Noise - 40uV
Signal To Noise Ratio - 124 dB
Output Impedance - 2.6 mOhms
Frequency Response - 10Hz to 50kHz (+0/-3dB)
Voltage Gain 26dB

2) and this one (Nagra Classic Amp) for the coaxial Midrange/tweeter unit

Power 100 W RMS per channel into 8 Ω Sensitivity 1 V or 2 V RMS Bandwidth 10 Hz to 80 KHz, +0/-3 dB Crosstalk >70 dB Signal-to-noise ratio Typically 110 dB ASA “A” weighted THD+N < 0.05% Input impedance >100 KΩ Automatic start For input level >10 mV Protection, Overheating > +60° C (140° F) deactivates the amplifier DC speaker protection above ± 2.5 VDC

could do any harm to the speakers or could cause any other problem.

Thanks!

@Olli If the HF and LF crossover filters are completely separate in the speaker, by that I mean the low isnt commoned (it shouldnt be for biwiring), I dont see there being any issue at all.

Although the P252 would possibly be a more appropriate choice.

Is the CR1 similar? If so the impedance plot for the isnt too terrifying

 

[Olli]

Thanks! The Ce-1 should be similar to the CR-1. It has jumpers for the bi-wiring terminal, so I believe they are independent.

Would you go for the exact manufacturer‘s XO settings or extend them a bit? Should be gouverened by the passive XO anyways in the end, right?

 

[March Audio]

Thanks! The Ce-1 should be similar to the CR-1. It has jumpers for the bi-wiring terminal, so I believe they are independent.

Would you go for the exact manufacturer‘s XO settings or extend them a bit? Should be gouverened by the passive XO anyways in the end, right?

...mmmm..the passive XO is going to do its thing anyway, you just want to correct it to follow the ideal slope, phase and be flat within band. If you go wider the passive XO will filter it any way. I have not used audiolense, I use acourate, so I dont know its features for correcting passive speakers, but essentially you should be following the passive XO.

What you really want to do is bypass the internal XOs completely and do all the filtering in DSP.

 

[March Audio]

Actually the more I think about this the more I think you should perform an "overall" correction. Dont DSP filter into separate bands. Send the overall signal to both LF and HF amps and let Audiolense correct the overall measured response.

Also, does the gain of your existing amp match the 26dB of the Ncores? A small difference wont matter but.....

 

[March Audio]

@mitchco ?

 

[Olli]

good point. Not sure about the gain. Will check!

 

[Olli]

good point. Not sure about the gain. Will check!

Just found an interesting test on an aussie website!

https://www.avhub.com.au/product-reviews/hi-fi/nagra-classic-preampclassic-amp-review-test-512382

But no gain info yet...

 

[March Audio]

@Olli The NAGRA website says sensitivity is "1 or 2 V" So for 100watts that equals 29 or 23dB. So just be aware you will have to match levels.

Thanks for the link to that review. It was interesting but unfortunately very disappointing. Hate to be the bearer of bad news but the Nagra is very high distortion and relatively noisy.

The overall THD+ Noise is 0.02% or SINAD of 74dB. This is very poor performance. SNR is 92dB.

These are some measurements from my Hypex Ncore P252. THD+Noise is 0.0015% or SINAD of 96.5dB. SNR121dB

I would be tempted to put the nagra on the woofer where the distortion would be less obvious.....or dare I suggest (without trying to be too much of a schill) sell it and use 2 P252s.

 

[Soniclife]

So if I biamp the TADs I can have 2 channels for the subs below 70 Hz or so, and direct 4 more channels to the Bi amping terminals of the TADs.
There's one 18cm cone Woofer and one coaxial Midrange/tweeter unit; so I could time align these three drivers with Audiolense when bi amping.

What sort of measured problem with the TAD drivers are you seeing currently? To me it feels like you are going for all the complexity of a DIY active crossover solution, with minimal amounts of the gain from it.

 

[March Audio]

Omg, just noticed, is the price of the nagra pre and amp really $47000 AU?

 

[mitchco]

@Olli As @March Audio says, I would not create separate digital XO's for the woofer and mids as the passive XO is still in play. You could try it, but I am not sure what audible benefit will occur. I say that as Audiolense also applies an overall excess phase correction, even with the passive XO in place. There also might be some gain matching involved... It is easy enough the create correction filters for both scenarios and see what the simulation looks like and then AB listen...

 

[March Audio]

@mitchco sorry off topic but whilst you are here, what's your view of audiolense V acourate?

 

[mitchco]

@March Audio for full featured DSP software, both are best in class. If using digital XO capabilities, Audiolense gets the nod as the time alignment process is fully automatic, and repeatable, whereas in Acourate it is a manual process. Given Acourate's manual process, introduces variability in time aligning drivers, especially for subs, as the wavelengths are so long. OTOH, depending on one's needs. Acourate has more raw audio DSP functions whereas Audiolense has a more automated workflow for most common use cases, including digital XO and time alignment.

There are a few idiosyncrasies between the two wrt target designers and slightly different proprietary psychoacoustic filtering, but the end results are top notch for both. Audiolense can lower low frequency group delay and cancel the first major room reflection... While Audiolense has multiseat correction capabilities, I rarely use it as both Audiolense and Acourate use special psychoacoustic analysis algorithms that does not require multiple analysis measurements. Both low frequency correction and time alignment remain solid across a wide listening area based on a single analysis measurement. I show in my book how the low end correction and time alignment remain the same across a 6ft x 2ft listening area with 14 separate validation measurements.

Both can do partial correction to any frequency. If using full range correction, the number one mistake people make is over correction at high frequencies, which leads to harsh top end. If using constant directivity waveguides, in a digital XO scenario, as is the case for me, then some level of constant directivity horn equalisation is required. However, I leave the top octave alone and let the compression driver and waveguide combo do its thing and always sounds better to my ears. If using full range correction, it is best to reduce the amount of correction in the high frequencies so the correction is more like a sloping or tilting "tone control" so one can adjust the amount of high frequency energy coming at you to taste.

Hope that helps.

 

[Olli]

Omg, just noticed, is the price of the nagra pre and amp really $47000 AU?

Not sure what the price is in Australia, but it is an expensive Amp - retail price in Europe is roughly 15 k... mine is pre owned, but still no bargain.

But it looks nice

 

[Olli]

What sort of measured problem with the TAD drivers are you seeing currently? To me it feels like you are going for all the complexity of a DIY active crossover solution, with minimal amounts of the gain from it.

No problem honestly. But I am very intersted in optimising the time alignment of all drivers of my LS setup. But I don‘t want to rip the passive XO of my TADs apart (yet). So that’s why I thought about just using 2 amps with the biwiring terminals.

 

[Olli]

@Olli I would be tempted to put the nagra on the woofer where the distortion would be less obvious.....or dare I suggest (without trying to be too much of a schill) sell it and use 2 P252s.

That might indeed be the endgame, maybe even 3 252s since it is a 3 way speaker. I am just too hesitant to disconnect the passive XO of this almost new and rather expensive speaker with full warranty yet.

 

[March Audio]

@March Audio for full featured DSP software, both are best in class. If using digital XO capabilities, Audiolense gets the nod as the time alignment process is fully automatic, and repeatable, whereas in Acourate it is a manual process. Given Acourate's manual process, introduces variability in time aligning drivers, especially for subs, as the wavelengths are so long. OTOH, depending on one's needs. Acourate has more raw audio DSP functions whereas Audiolense has a more automated workflow for most common use cases, including digital XO and time alignment.

There are a few idiosyncrasies between the two wrt target designers and slightly different proprietary psychoacoustic filtering, but the end results are top notch for both. Audiolense can lower low frequency group delay and cancel the first major room reflection... While Audiolense has multiseat correction capabilities, I rarely use it as both Audiolense and Acourate use special psychoacoustic analysis algorithms that does not require multiple analysis measurements. Both low frequency correction and time alignment remain solid across a wide listening area based on a single analysis measurement. I show in my book how the low end correction and time alignment remain the same across a 6ft x 2ft listening area with 14 separate validation measurements.

Both can do partial correction to any frequency. If using full range correction, the number one mistake people make is over correction at high frequencies, which leads to harsh top end. If using constant directivity waveguides, in a digital XO scenario, as is the case for me, then some level of constant directivity horn equalisation is required. However, I leave the top octave alone and let the compression driver and waveguide combo do its thing and always sounds better to my ears. If using full range correction, it is best to reduce the amount of correction in the high frequencies so the correction is more like a sloping or tilting "tone control" so one can adjust the amount of high frequency energy coming at you to taste.

Hope that helps.

Thanks for the detailed reply @mitchco . Two more questions. I use acourate convolver for the replay. Main reason for this is that for home theatre I take the front 3 channels output from my AV amp via Motu interface. So using the "sound card" input instead of ASIO.
Does Audiolense have a similar facility to directly input from an A to D?

Second question, can audiolense create low latency minimum phase filters? Again for theatre use.